Benefits of and Safety Concerns Associated with Drug-eluting Coronary Stents

Causes of ST

The occurrence of ST remains largely unpredictable and no specific causative factor has been identified. However, it is possible to identify those patients at increased risk (Box 3).

Device Factors

Device-related factors that have been implicated in precipitating ST include those discussed in the following sections.

Impaired Endothelialization by Antiproliferative Drugs The antiproliferative properties of DES impair and/or delay endothelialization such that blood is exposed to thrombogenic stent struts, potentially precipitating ST. Scanning electron microscopy in animal models has shown that at 1 month postimplantation, endothelial coverage of stent struts varies among different stents, with DESs having a greater total area of uncovered stent strut than BMSs (Figure 3).^[57] Similar findings have also been shown in recent human studies using optical coherence tomography to assess endothelial stent strut coverage between different DESs at 9-months follow-up. Guagliumi et al. reported significantly reduced frequencies of uncovered stent struts with zotarolimus-eluting stent (ZES) compared with SESs in both overlapping (0.02 vs 5.8%) and non-overlapping sites (0.01 vs 6.0%).^[58] Kim et al. have recently demonstrated similar results,^[59] while Barlis et al. demonstrated a higher rate of near-complete (>95%) strut coverage in a stent with a biodegradable polymer when compared with that with a durable polymer (SES; 89.3 vs 63.3%; p = 0.03).^[60]

Polymers A hypersensitivity reaction to the presence of a permanent polymer may be a causative factor in ST.^[18,61] Furthermore, histopathological studies have shown that these durable polymers can cause localized vascular inflammation, hyper-eosinophilia, thrombogenic reactions and apoptosis of smooth muscle cells, all of which may precipitate ST.^[62–64]

Specifically, the non-erodable polymers poly(ethylene co-vinyl acetate) and poly(N-butyl methacrylate) found on the first generation Cypher SES have been shown to exclusively induce granulomatous and hypersensitivity reactions in animal models and humans. Similarly the first-generation TAXUS PES stent has a durable poly(styrene-b-isobutylene-b-styrene) polymer, which is associated with medial necrosis, positive remodeling and excessive fibrin deposition, which are all likely to be cytotoxic effects of paclitaxel.^[65]

The concerns with these first-generation stents have led to the development of DESs that have novel stent coatings, which are more biocompatible or bioabsorbable. In addition DESs that are polymer free have been developed, together with polymeric and metallic stents that are completely bioabsorbable. These are discussed further in the 'Five-year view' section of this article.

The second-generation Endeavor® (Medtronic, MN, USA) ZES uses a phosphorylcholine polymer, which although not biodegradable, is biostatic and a natural component of the cell membrane. The polymer is biomimetic and biocompatible, and studies have shown it causes less inflammation compared with the polymers in the Cypher SES stent. The release kinetics of zotarolimus enable complete drug release within 1 month of stent deployment. Clinical data have shown conflicting results with ZESs compared with other DESs for early and late ST; however, a benefit has been seen with ZESs in the reduction of very late ST.

At 9-months follow-up in the SORT-OUT III study, which randomized 2333 patients between ZESs (1162) and SESs (1171), ZESs had higher rates of cardiac death (p = 0.14), MI (p = 0.03), clinically significant restenosis (p ≤ 0.0001) and early/late ST (p = 0.02).^[201] The Comparison of Sirolimus- and Paclitaxel-Eluting Stents vs Zotarolimus-Eluting Stents in Real World Practice (ZEST) trial randomized 2640 patients to PESs, SESs or ZESs; the primary end point of major adverse cardiovascular events (MACE) was similar between SESs and ZESs (p = 0.25), and significantly higher with PESs. On the other hand, ST at 1 year was similar between all three stents, with rates of 0.5, 0.0 and 0.7% for ZESs, SESs and PESs, respectively.^[66] In the ENDEAVOR IV study, which randomized 1548 patients to ZESs or PES there were comparable rates of MACE at 2-years follow-up. The rates of early and late definite/probable ST were higher with ZESs (0.8 vs 0.1%; p = 0.12), but very late ST was lower (0.1 vs 0.8%; p = 0.07).^[202] Similarly, in a pooled analysis of over 2500 patients, those treated with ZESs had a lower rate of ST out to 5 years compared with those treated with the DRIVER® (Medtronic, MN, USA) BMS (0.8 vs 1.7%; p = 0.069).^[203] Finally, recently pooled data from all DES trials have shown lower rates of very late ST with ZESs compared with other DESs.^[203]

The Xience V™ everolimus-eluting stent (ABBOTT Vascular, CA, USA) has cobalt chromium struts and a durable biocompatible polymer consisting of acrylic and fluorinated polymers that are 6–8 µm thick. The stent releases approximately 80% of the everolimus in the first month, and nearly 100% of it by 4 months. Studies have shown it to have a greater efficacy and lower overall rates of ST out to 3-years follow-up (0.9 vs 2.7%; p = 0.27) when compared with PESs.^[67,68]

Other Device-related Factors Other factors that have been suggested to influence ST include stent design, with lower platelet activation noted in those with closed- versus open-cell stents^[69] the composition of the metal struts;^[70] the deployment of stents at sites of necrotic core,^[61] and characteristics of the stent surface, with electromechanically polished stainless steel associated with a less thrombogenic surface.^[71]

Procedural Factors

Stent Expansion Major postprocedural predictors of ST with both BMS and DES are minimal stent cross-sectional area, and stent under-expansion. Stent under-expansion is assessed by calculating the ratio between the stent cross-sectional area at the site of the lesion and the balloon cross-sectional area, which is in turn calculated according to the compliance chart by using the same atmosphere as stent deployment.^[72–74] It has been suggested that these mechanical factors, which can result in abnormal shear stresses, are most important for the development of early ST. Conversely, biological factors such as incomplete stent apposition and positive remodelling have a greater influence on the development of late/very late ST. Fuji et al.^[75] and Alfonso et al.^[76] have both reported that lesions leading to ST after successful PCI with DESs more often have evidence of stent under-expansion, smaller minimal surface areas and a residual edge stenosis. Recently, Liu et al. reaffirmed these findings, and also demonstrated that those patients presenting with ST have stent under-expansion that is more severe, diffuse and proximal when compared with those presenting with just restenosis.^[77] The risk of ST due to stent undersizing has been shown in the recently published Dutch Stent Registry of 21,900 patients to be greater in those with stable angina (OR: 57.7; 95% CI: 6.26–23.9; p = 0.0003) compared with those presenting with acute coronary syndrome (ACS; OR: 12.28; 95% CI: 4.72–31.93; p < 0.0001).^[74] Stent expansion can be improved at the time of deployment by the use of prolonged stent balloon inflation,^[78] and by the use of intravascular ultrasound (IVUS), both presenting to adequately size the stent and poststenting to evaluate deployment.

Incomplete Stent Apposition Incomplete stent apposition (ISA) represents a separation of at least one strut from the intimal surface of the vessel, with evidence of blood behind the strut. It can be acute if detected at the time of the procedure, or late if detected on follow-up IVUS. Acute ISA can resolve itself, or can be treated with balloon angioplasty at the time of PCI. Late ISA can be persistent (present postprocedure and at follow-up) or acquired, if only detected on follow-up.^[79] ISA is associated with an increased risk of ST. An IVUS study in 13 patients with very late ST demonstrated frequent ISA (77 vs 12%, p < 0.001) and a larger maximum incomplete stent apposition area in patients with very late ST compared with controls who comprised of a cohort of patients from the Sirolimus Eluting Versus Paclitaxel Eluting Stents for Coronary Revascularization (SIRTAX) study treated with DES who returned for angiographic follow-up (8.3 vs 4.0 mm²; p = 0.03).^[80] Most recently, a meta-analysis has shown that the risk of late-acquired ISA is significantly higher for DES compared with BMS (OR: 4.36; 95% CI: 1.74–10.94). Similarly, the risk of late/very late ST was significantly higher in those with ISA compared with those with ISA (OR: 6.51; 95% CI: 1.34–34.91).^[81]

This difference in malapposition between DES and BMS may be due to the effect of the antiproliferative drug on the vessel wall, causing positive remodeling, or a result of the decrease in plaque volume behind the stent struts.^[72] It has also been demonstrated using optical coherence tomography in patients pre-PCI, and at 9-months follow-up that lesions with plaque rupture, thrombus, lipid-rich plaque and thin-capped fibroatheroma have a greater incidence of ISA than those without these features at baseline (83 vs 30%; p < 0.001).^[82]

However, it is not entirely clear exactly how ISA leads to ST, but it may be the result of chronic inflammation and delayed healing, causing tissue necrosis and erosion around the stent.^[61] The link between inflammation, ISA and ST has been reaffirmed by a recent histopathological study of ST eosinophil counts, which demonstrated that not only is very late ST associated with a greater degree of inflammation than other types of ST (early, late and BMS), but eosinophil counts appear to correlate with the degree of ISA.^[83] Finally, ISA may serve as a trigger for thrombosis by allowing fibrin and platelet deposition behind stent struts.^[84]

Lesion Factors

Different lesion characteristics, such as long lesion length, small vessel size and high lesion complexity have all been suggested as having an influence on the risk of ST; however, it must be appreciated that few studies of individual lesion characteristics are sufficiently powered to establish a confirmatory link. Many of these lesion characteristics are considered as 'off-label' indications for DESs, and therefore as previously discussed, DESs may not confer a higher risk of ST compared with BMSs. Furthermore, data from the ARRIVE registry have indicated that the multivariate predictors of ST change with follow-up time, such that anatomical lesion characteristics; for example, lesion length over 28 mm, multiple-stent deployment, calcification and small vessel diameter all influence early/late ST, but not very late ST. On the other hand, stenting a CTO and biological factors including previous brachytherapy only influence very late ST, and not early/late ST.^[47]

The association of these lesion characteristics with ST stems from a combination of the procedural factors listed earlier; for example, lesion length is an independent predictor of late ISA,^[85] and incomplete stent expansion is more likely in complex angulated anatomy such as found in a bifurcation lesion. Other factors include the increased number of stents that may be needed, which increases the surface area of blood vessel exposed to the drug and thrombogenic stent struts.

Long Lesions/Small Vessels Stent length has been shown to be an independent predictor of acute and late ST;^[45,47] however, recent registry data have also suggested that rates of ST are lower for DES compared with BMS for stent lengths of 18 mm or more (1.5 vs 2.9%) and stent diameters of 3.0 mm or more (1.4 vs 1.7%).^[86]

Complex Lesions Data from both the ARRIVE registry and SYNTAX study indicate that lesion complexity increases the risk of ST. The rates of ST at 1 year in the ARRIVE registry for 'simple lesions' (single-vessel, single-stent lesions) and 'expanded-use lesions' (comprising bifurcations 7.7%, CTO 2.1%, left-main stenting 2.2%) were 0.9 and 2.2%, respectively. These compare with the higher 3.3% rate of ST at 1 year in the SYNTAX trial of complex lesions (comprising bifurcation lesions: 72.4%; CTOs: 24.2%; left-main stenting: 34.6%) using the same TAXUS Express^[2] PES stent.^[41,47,87]

Bifurcation Lesions Numerous registries have reported that bifurcation lesions are an independent predictor of ST.^[26,74] Studies of bifurcation lesions have demonstrated that they have higher rates of ST when compared with nonbifurcation lesions treated with the same DES (p = not significant).^[88,204] Similarly, studies have shown that treatment of bifurcation lesions using a complex stenting strategy does not increase the risk of ST when compared with using a single-stent strategy. At 6 months follow-up Colombo et al. reported ST rates of 1.1 and 1.7% (p = 0.62), while at 1-year follow-up, Ferenc et al. reported rates of 1.0 and 2.0% (p = 1.0), and at 3-years follow-up Sjörgen et al. reported rates of 2.5 and 1.0% (p = 0.44) for respective simple and complex (two-stent) bifurcation stenting strategies.^[89,90,205] Furthermore, no significant difference in ST has been shown between different complex stenting strategies; Erglis et al. reported a rate of ST of 1.4 and 1.9% for patients randomized to crush or culotte stenting, respectively (p = 0.73).^[91] The use of final kissing-balloon inflation, which improves both stent expansion and strut apposition, can influence the rate of ST. This was demonstrated in the Coronary Bifurcations: Application of the Crushing Technique Using Sirolimus-Eluting Stents (CACTUS) study, which reported a higher rate of ST in those patients who did not receive a final kissing-balloon inflation (0.9 vs 6.5%; p = 0.06).^[90]

CTOs The ARRIVE registry demonstrated that PCI for a CTO is a risk factor for very late ST. In previous dedicated studies of CTOs, few ST events have been reported, which may be secondary to the small number of patients recruited in these studies, the short follow-up, or under-reporting, because these patients are usually well collateralized such that vessel occlusion may remain asymptomatic. For example Valenti et al. reported a rate of re-occlusion in successfully opened CTO lesions of 11.2% in the 282 patients who returned for angiographic follow-up at 8 months; however, only three patients experienced an MI.^[92] The rate of ST was not reported. In the longest follow-up study of CTO patients, Shen et al. reported no significant difference in ST amongst CTO patients treated with SES or BMS up to 5-years follow-up.^[93]

Patient Factors

Clinical Presentation: ACS/ST-elevation MI The prognosis of patients with ACS or ST-elevation MI (STEMI) has been improved with the use of invasive reperfusion therapy;^[94,95] however, there are concerns that this may be associated with higher rates of ST.^[27,96,97] This stems from the potential trapping of thrombus behind stent struts, which can potentially increase the risk of ISA. In addition, owing to plaque rupture, stent struts can protrude into underlying necrotic core. Recent histopathological data also suggest this increased risk of ST is secondary to the greater degree of delayed arterial healing (e.g., greater incomplete stent strut endothelialization and persistent fibrin deposition) shown at the culprit site in STEMI patients, compared with patients treated for stable angina.^[98] Finally, it is difficult to predict how compliant patients will be to DAPT given the emergent nature of the procedure, and the minimal time operators have to discuss these issues with patients before their PCI. The Prospective Registry Evaluating Myocardial Infarction: Events and Recovery (PREMIER) study worryingly showed that 13.6% of ACS/STEMI patients had discontinued clopidogrel within 30 days of their PCI, leading to a significantly increased risk of mortality in the following 11 months.^[99]

Clinical data has reaffirmed concerns regarding the higher rates of ST in patients with ACS/STEMI. Kukreja et al. recently reported the results of a large registry of 5816 patients treated with BMSs, SESs and PESs. Results showed that at a median of 3.8 years follow-up the rate of definite ST was 2.5% in patients with ACS (STEMI, non-STEMI and unstable angina) and 1.0% in stable patients (Figure 4). ACS was identified as a risk factor for early/late ST with both BMS and DES, whilst the risk of very late ST was only increased in ACS patients treated with DES. Of note, the risk for death from ST was significantly higher in those with stable angina, which may reflect the benefits of ischemic preconditioning in these unstable patients.^[100]

Data on ST in primary PCI are limited by the short follow-up of most studies. The largest randomized primary PCI trial to date, Harmonizing Outcomes with Revascularization and Stents in Acute Myocardial Infarction (HORIZONS-AMI), has outcome data to 12 months, and recently reported no significant difference between 3006 STEMI patients treated with PES or BMS in terms of mortality (3.5 vs 3.5%; p = 0.98) or ST (3.2 vs 3.4%; p = 0.77).^[101] Similarly, a meta-analysis of 2786 patients by Kastrati et al. reported no significant difference in terms of death, MI or ST between BMS or DES in primary PCI; however, follow-up did not extend beyond 2 years.^[102] A much larger meta-analysis of 33,873 STEMI patients by Brar et al. reported similar findings, again at a maximum of 2 years follow-up.^[103] Of note, among 7352 patients from 13 randomized, clinical trials, the incidence of ST at 1 year was 2.7 and 2.6% for DES and BMS, respectively (p = 0.81). Only four trials subsequently reported ST at 2 years, which was again comparable with both types of stent (p = 0.52).

Long-term data have recently been presented on two STEMI trials that show encouraging results. The Trial to Assess the Use of the Cypher Stent in Acute Myocardial Infarction Treated with Balloon Angioplasty (TYPHOON) study randomized 715 STEMI patients to PCI with either a BMS or SES stent. The 1-year results showed comparable death, MI and ST. The trial by design was closed at 1 year, but in view of the concerns regarding very late ST, the trial was 'reopened' to assess 4-year follow-up. At 4 years, the rate of death, cardiac death and MI was comparable between both groups, whilst TLR/TVR was significantly less in the SES group. The rate of definite ST was similar between both groups (SES 3.6% vs BMS 4.0%; p = 0.82), but very late ST was higher with SES (2.0 vs 0.8%).^[206] Although these results are encouraging, the extensive list of exclusion criteria (massive thrombus in the infarct-related vessel, bifurcation lesions, calcification, multiple lesions and so on),^[104] which resulted in only 35% of those patients screened being enrolled, must be considered when interpreting the results, together with complete follow-up being available in only 70% of patients.

The 5-year outcomes from the single-center Single High-Dose Bolus Tirofiban and Sirolimus-Eluting Stent vs Abciximab and Bare-Metal Stent in Myocardial Infarction (STRATEGY) trial showed a significant reduction in the primary end point (a composite of death, MI, stroke and TVR) in patients with STEMI randomized to treatment with either DES and tirofiban or BMS and abciximab (30 vs 51%; p = 0.006). The rate of overall ST was comparable between both groups (7 vs 8%; p = 0.78).^[207]

In summary, although registry data suggest an increased risk of ST with DES in patients with ACS, this has not been confirmed in randomized trials or meta-analyses. Long-term data, however, are awaited.

Diabetes Mellitus Diabetes has been shown to be an independent risk factor for ST.^[74] This increased risk is likely to be related to the diabetic proinflammatory and prothrombotic state, which is a consequence of insulin resistance and hyperglycemia-induced changes to platelet function and coagulation.^[105] In addition, this increased risk is associated with the multiple comorbidities frequently seen in diabetics,^[106] together with their diffuse and aggressive atherosclerosis, which often leads to long lesions in small-caliber arteries with a greater plaque burden, all of which may lead to less optimal procedural results.^[107–109] Finally, diabetics appear to be greater nonresponders to antiplatelet therapy.^[110]

The specific assessment of the safety of DES in diabetic patients has yet to be performed in a randomized clinical trial; however, the matched-cohort Évaluation Coût/Efficacité du Stent Actif au Sirolimus Chez les Patients Diabétiques et non Diabétiques (EVASTENT) registry assessed the frequency and causes of ST in 1731 diabetic and nondiabetic patients treated with SES. At 12-months follow-up the rate of all-cause mortality and cardiovascular mortality was significantly higher in the diabetic group. The rate of ST was 1.8-times higher in diabetics than in nondiabetic patients (3.2 vs 1.7%; log rank p < 0.03). In a multivariate analysis, in addition to the interruption of antithrombotic treatment, independent ST predictors were previous stroke, renal failure, lower ejection fraction, calcified lesion, length stented and insulin-requiring diabetes.^[106]

Duration of DAPT The introduction of thienopyridines was instrumental in the reduction of thrombotic events post-PCI,^[111,112] but to this day the optimal duration of DAPT continues to remain an issue of contention. It has been shown repeatedly that the early cessation (<1 year) of DAPT is one of the most significant independent predictors of ST.^{[26,74,99,113]} Commonly cited reasons for this early cessation include poor patient compliance, surgery, bleeding complications, poor patient education, allergy to clopidogrel and cost.^[4,99]

The use of the word 'early' is highly contentious because the guidelines over the duration of clopidogrel therapy post-PCI are not based on actual scientific evidence, and appear to be somewhat arbitrary.^[114] Of note, no long-term issues with regard to ST have been observed out to 5 years follow-up in the pivotal RAVEL and SIRIUS studies where clopidogrel was given for 2 and 3 months, respectively.^[53,115] The TAXUS I study was the first to recommend 6 months of clopidogrel, although the basis for the change from 2 or 3 months is not apparent.^[116] Only recently have guidelines from the American Heart Association/American College of Cardiology advised that the duration of clopidogrel be extended to 12 months post-PCI.^[117] This extension, together with suggestions that DAPT should be continued for life, have been debated in the literature,^[118,119] in particular because approximately 99% of patients who discontinue DAPT never experience a ST event.^[6]

ST can occur in patients still taking DAPT, as illustrated in Figure 5.^[113] Other studies have confirmed similar findings; for example, Daemen et al. reported that 87% of patients with early, and 23% of patients with late ST in the large Bern–Rotterdam cohort were still taking DAPT at the time of the ST. The absence of clopidogrel was not demonstrated to be an independent predictor of ST in this population.^[4] Conversely, in the ARRIVE registry, although Lasala et al. reported that 71.4% of patients with early and 23.1% of patients with very late ST were still taking DAPT at the time of the event, premature discontinuation of DAPT was still shown to be a strong independent predictor of ST.^[47]

Studies have demonstrated that the discontinuation of clopidogrel is only a major independent predictor of ST in the first 6 months following PCI, and not beyond. The median time interval for a ST event following the discontinuation of clopidogrel has been shown to be 9 days (interquartile range [IQR]: 5.5–22.5) within the first 6 months of the PCI, compared with 104.3 days (IQR: 7.4–294.8) for the period afterwards.^[113,120]

Long duration of clopidogrel therapy exposes patients to the risks of bleeding. The Clopidogrel in Unstable Angina to Prevent Recurrent Events (CURE) trial reported higher rates of major bleeding (3.7 vs 2.4%) and minor bleeding (5.1 vs 2.4%) for treatment with DAPT compared with aspirin on its own. However, in certain subgroups of patients, this increased risk of bleeding was outweighed by the benefit in terms of cardiovascular death, MI and stroke.^[121]

Irrespective of timing, the discontinuation of clopidogrel has been shown in diabetic patients with coronary artery disease to be both proinflammatory and prothrombotic.^[122] Furthermore, in nondiabetics recent data from a small multicenter study of 98 patients treated with 12 months of clopidogrel post-PCI, have shown an increase in the proinflammatory cytokine soluble CD40L 4 weeks after the discontinuation of clopidogrel. This may reflect an increase in proinflammatory state; however, a corresponding reduction in high-sensitivity C-reactive protein makes this less likely.^[123]

The financial implications of long-term clopidogrel treatment may not be as prominent an issue once the drug becomes generic.

Ongoing trials, for example Intracoronary Stenting and Antithrombotic Regimen: Safety and Efficacy of a 6-month DAT After Drug-Eluting Etenting (ISAR-SAFE), Correlation of Clopidogrel Therapy Discontinuation in Real-World Patients Treated with Drug-Eluting Stent Implantation and Late Coronary Arterial Thrombotic Events (REAL-LATE) and Clopidogrel Use and Long-Term Safety After Drug-Eluting Stents Implantation (ZEST-LATE), will hopefully provide more definitive answers on the ideal duration of clopidogrel therapy. In particular, the ISAR-SAFE study is randomizing 6000 patients treated with DES to either 6 or 12 months of DAPT. The primary end point is a composite of death, MI, ST, stroke and TIMI major bleeding at 9 months, and results should be available in late 2011 (clinical trial number: NCT00661206).^[124]

Clopidogrel 'Resistance'/'Nonresponders' Resistance to aspirin and/or clopidogrel is emerging as an important clinical entity that has the potential to expose patients to an increased risk of adverse cardiac events,^[125] and in particular ST.^[126,127] The reported incidence of clopidogrel nonresponsiveness or resistance varies from 5 to 44%,^[128,129] depending on the loading dose of clopidogrel administered, the time between dosing and the measurement of platelet function, and the method used in determining nonresponsiveness.

Currently, the mechanism of resistance remains incompletely defined, but it is likely to occur through a combination of clinical, cellular and genetic factors, together with drug interactions as listed in Box 4.^[130] Of note, clopidogrel is a prodrug, such that the two-step process of converting clopidogrel into its active metabolite involves several cytochrome P450 (CYP) enzymes whose activity varies considerably between individuals. Specifically, the loss of function associated with the CYP2C19*2 polymorphism is associated with inactivation of the enzyme, and impaired metabolism of clopidogrel, leading to poor response to the drug.^[131]

There have been recent developments in near-patient tests to rapidly and accurately assess platelet function. In particular, the reactive platelet response to ADP in those patients with ACS on clopidogrel, assessed using the point-of-care assay VerifyNow® (Accumetrics, CA, USA), has been shown to be able to identify those at risk of cardiovascular death and nonfatal MI at 12 months follow-up.^[132]

The treatment of clopidogrel resistance has not been fully established, and in view of potential fatal consequences is a major clinical problem. Simple measures include ensuring adequate patient compliance and evaluating possible drug interactions. Additional strategies are discussed below.

Increased Clopidogrel Dosing Improvements in platelet inhibition and subsequent clinical outcomes have been demonstrated by using a 600-mg or more loading dose of clopidogrel, compared with a standard 300-mg dose, or by using a higher maintenance dose (150 mg/day).^[133–135] Recent observational data also indicate that a combined loading dose of 600 mg followed by a maintenance dose of 150 mg/day improves clinical outcomes without significantly increasing bleeding events.^[136] The recently presented Clopidogrel Optimal Loading Dose Usage to Reduce Recurrent Events/Optimal Anti-platelet Strategy for Interventions (CURRENT OASIS 7) trial demonstrated that doubling the loading and maintenance dose of clopidogrel (for 7 days) improves rates of cardiovascular mortality, MI and stroke at 30 days in patients undergoing PCI. The study also reported a significant 22% reduction in ST in the group that received high-dose clopidogrel compared with the group that received the standard dose. Although the high-dose group had more major bleeding, there was no increase in intracerebral or fatal bleeds. Patients who did not undergo PCI sustained no benefit from the high-dose regime.^[137]

Triple-antiplatelet Therapy During PCI, the use of glycoprotein IIb/IIIa inhibitors has been shown to improve clinical outcomes in patients who are poor responders to aspirin/clopidogrel.^[138,139] During the maintenance phase, triple therapy can be achieved with the addition of cilostazol, a phosphodiesterase III inhibitor. This has been shown among diabetics to enhance P2Y₁₂ inhibition, without an increased risk of bleeding.^[140]

Alternative P2Y₁₂ Receptor Antagonists Ticlopidine has been largely replaced by clopidogrel owing to an improved safety profile; however, it may improve platelet inhibition in some patients who are poor responders to clopidogrel.^[141] The newer antiplatelet agents prasugrel and ticagrelor are discussed in the 'Five-year view' section of this article.

Special Conditions: Noncardiac Surgery after DES In the perioperative period, the increased risk of ST is related to the combination of the cessation of DAPT, nonendothelialization of the stent struts and a hypercoagulable state following surgery.^[142] Studies have shown that the risk is greatest when surgery is performed within 1–2 months of the PCI; however, a persistent risk remains for surgery performed 2–3 years after the index PCI. Furthermore, recent results indicate that the continuation of DAPT in the perioperative period may not offer universal protection from ST.^[143]

Some have considered heparin as a suitable substitute perioperatively, but studies have shown the use of intravenous heparin up until the time of surgery to be a univariate predictor of ST (p = 0.006).^[144] This may be owing to platelet activation, which, together with the hypercoagulable state, can precipitate ST. Interestingly, however, only one patient experienced an acute ST among the 721 patients with stable/unstable angina, or acute MI treated with a stent coated with covalently bonded heparin in the Belgain Netherlands Stent (BENESTENT) pilot, BENESTENT II and Primary Angioplasty in Myocardial Infarction (PAMI) pilot.^[145–148] Unfortunately, this technology fell out of favor with the development of newer BMSs and DESs; however, there is now scope for it to return with the current concerns over ST, particularly if the same results can be achieved when heparin is combined with an antiproliferative drug coating.

Glycoprotein IIb/IIIa inhibitors may play a role, but the resulting increased bleeding risk from surgery is likely to prevent their acceptance amongst most surgeons. If a surgeon is agreeable, tirofiban is considered the most appropriate agent to use as it provides the most rapid restoration of platelet aggregation after discontinuation of the drug infusion.^[149] Reports suggest that tirofiban can be discontinued 3–6 h prior to the procedure, the optimal time being dependent more on the type of surgery being performed and the magnitude of complications that bleeding might cause than the overall risk of bleeding.^[150] Interestingly, a recent study has demonstrated that tirofiban can reduce the extent of myocardial damage in patients who are poor responders to DAPT.^[139]

Recent guidelines suggest that elective noncardiac surgery should be delayed for 1 year following PCI with DES to reduce the risk of ST. Furthermore, clopidogrel should be stopped for as short a time as possible, and if possible aspirin should be continued throughout. Obviously, the nature of the surgery and the preference of the surgeon may result in a deviation from these recommendations.

Treatment of ST & Prognosis

The most common presentation of ST is acute MI, and therefore according to current treatment guidelines patients should be treated with primary PCI in a similar manner to those patients presenting with STEMI due to a rupture atheromatous plaque. Studies have shown that despite the prompt restoration of blood flow through the occluded stent with primary PCI, these patients tend to experience substantial myocardial damage. Furthermore, ST is associated with a poorer outcome, which is independent of underlying risk factors. In a series of 2464 consecutive STEMI patients treated with primary PCI, the 3% of patients presenting with ST were found to have a higher mortality, and a significantly higher risk of re-infarction, restenosis and TVR at 6 months.^[151] Similarly, among a cohort of 431 patients with ST treated with primary PCI, van Werkum et al. reported a recurrent ST rate of 20% within 3 years. Clinical outcomes were not affected by whether the ST was early or late, or due to a BMS or DES. In 49.7% of patients, the ST was treated with the implantation of an additional stent; however, this was shown to be an independent predictor of death/recurrent ST event (HR: 1.73; 95% CI: 1.14–2.61; p < 0.001) at follow-up. This important finding indicates the need for randomized trials to assess the optimal treatment of ST.^[144]

Expert Rev Cardiovasc Ther. 2010;8(3):449-470. © 2010  Expert Reviews Ltd.

Cite this: Benefits of and Safety Concerns Associated with Drug-eluting Coronary Stents - Medscape - Mar 01, 2010.